Influence of calcium nitrate timing on the structural and textural characteristics of mesoporous SiO₂-CaO nanoparticles

Influence of calcium nitrate timing on the structural and textural characteristics of mesoporous SiO 2 -CaO nanoparticles Onat Bas ¸ak a,* , Fatih Kurtuldu a,* , Jan Ilavský b , Martina Vit´ azkov´ a a , Ana M. Beltr´ an c , Francisco Mu˜ noz d , Yolanda Castro d , Martin Mich´ alek a , Maria Chromˇ cíkov´ a a,e a FunGlass, A. Dubˇ cek University of Trenˇ cín, ˇ Studentsk´ a 2 911 50 Trenˇ cín, Slovakia b X-ray Science Division, Advanced Photon Source, Argonne National Laboratory, Argonne 60439, IL, USA c Departamento de Ingeniería y Ciencia de los Materiales y del Transporte, Escuela Polit´ ecnica Superior, Universidad de Sevilla, Sevilla, 41011, Spain d Institute of Ceramics and Glass (CSIC), Kelsen 5, Madrid 28049, Spain e VILA – Joint Glass Centre of the IIC SAS, TnUAD, FChPT STU, ˇ Studentsk´ a 2 911 50 Trenˇ cín, Slovakia ARTICLE INFO Keywords: Mesoporous bioactive nanoparticles Microemulsion assisted sol-gel synthesis Calcium nitrate tetrahydrate ABSTRACT Mesoporous bioactive glass nanoparticles (MBGNPs) are promising materials for drug delivery due to their high pore volume and specific surface area. This study investigates how the timing of calcium nitrate addition affects the structural and textural characteristics of MBGNPs synthesized via a microemulsion-assisted sol-gel method. Delayed calcium nitrate addition reduced CaO incorporation from 14.2 to 9.5 mol% and increased particle size from 178 ±51 nm to 256 ±30 nm. The specific surface area values increased with the delayed addition of calcium nitrate, as observed through BET and USAXS/SAXS measurements. The proportion of Q Si n units slightly changed, but no cytotoxicity was observed in osteoblast-like cells. These findings provide valuable insights into optimizing MBGNP synthesis for biomedical applications. 1. Introduction Bioactive glasses (BGs) have been extensively studied for biomedical applications since their development in the late 1960s due to their remarkable ability to bond with bone [1,2]. This unique property arises from the formation of a hydroxyapatite layer on their surface through dissolution in contact with physiological fluids, which mimics the mineral composition of bone. BG dissolution can be tailored by doping therapeutic ions and modifying chemical composition, which can enhance biological functions such as angiogenesis, osteogenesis, and antibacterial activity [3]. These properties have expanded the BG applications in wound healing and soft tissue engineering [4]. Additionally, the fabrication method, melting or sol-gel, notably influences the ion release behavior of BGs [5]. Melt quenching was the first method for fabricating BGs [6]. However, it has certain limitations, including the use of high processing temperatures and a narrow composition range to maintain bioactivity [4]. For instance, melt-derived glasses can incorporate only up to 60 mol % SiO 2 , whereas sol-gel synthesis allows the fabrication of glasses with up to 90 mol% SiO 2 at lower temperatures, while still preserving bioactivity [7]. This method involves hydrolysis and condensation reactions of precursors (alkoxydes or salts) to obtain and interconnected inorganic network at room temperature and ambient pressure using, for example, tetraethyl orthosilicate (TEOS) as the precursor [8]. Adjusting the catalyst and precursors concentration can modify the particle size and morphology, thereby influencing bioactivity [9]. These advantages make the sol-gel method ideal for synthesizing advanced materials such as porous scaffolds, hybrid BGs, and mesoporous bioactive glass nanoparticles (MBGNPs) [10]. MBGNPs are promising hosts for various drugs, genes, and growth factors due to their superior surface area, larger pore volume compared to conventional BGs [11–13]. Additionally, MBGNPs exhibit higher purity and homogeneity than melt-derived glasses and can be further modified due to the presence of silanol groups [14]. In sol-gel synthesis, calcium nitrate is widely used as a calcium precursor due to its solubility in water and low cost [15]. However, its incorporation into the silica network presents significant challenges. Calcium nitrates are only incorporated into the glass structure at temperatures above 450 ◦C, whereas silica network formation begins from room temperature [16]. This mismatch in reaction conditions often * Corresponding authors. E-mail addresses: [email protected] (O. Bas ¸ak), [email protected] (F. Kurtuldu). Contents lists available at ScienceDirect Open Ceramics journal homepage: www.sciencedirect.com/journal/open-ceramics https://doi.org/10.1016/j.oceram.2025.100807 Received 14 April 2025; Received in revised form 15 May 2025; Accepted 3 June 2025 Open Ceramics 23 (2025) 100807 Available online 12 June 2025 2666-5395/© 2025 The Author(s). Published by Elsevier Ltd on behalf of European Ceramic Society. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ). leads to the formation of calcium-rich regions rather than a homogeneous distribution within the silicate network [17]. Furthermore, discrepancies between nominal and experimental compositions have been frequently reported in studies using calcium nitrate as a precursor [18–20]. These discrepancies are primarily attributed to the washing and centrifugation processes, which are essential for removing unreacted precursors and surfactants, but can also result in the loss of calcium ions. While these steps are necessary to maintain particle dispersion and prevent unwanted side reactions, they exacerbate the gap between the desired and achieved compositions. The micro-emulsion assisted sol-gel method is one of the most common and facile approaches for the synthesis of MBGNPs, offering the advantage of preventing particle agglomeration during synthesis [8,21]. In this method, hexadecyltrimethylammonium bromide (CTAB) acts as a surfactant to facilitate the formation of mesoporous structures. Under basic conditions, CTAB, combined with ethyl acetate (EA), forms oil-in-water (O/W) microemulsion droplets, which serve as confined reaction sites for the synthesis process [22]. Within these droplets, the hydrolysis and condensation of TEOS occur in the presence of aqueous ammonia, which functions as both a catalyst and a regulator of textural properties [23]. The concentrations of ammonia and CTAB play a critical role in controlling particle size, mesopore structure, and morphology. For example, Liang et al. [24] observed that varying ammonia concentrations led to different pore types (e.g., worm-like, radial, and lamellar) and an increase in particle size from 28 to 255 nm. Similarly, Li et al. [25] reported differences from spheres to rod-shaped particles with increasing CTAB concentration. These findings highlight the tunability of MBGNP properties through precise control of synthesis parameters. Several studies have shown that the timing of calcium nitrate addition significantly influences the textural and structural properties of MBGNPs. For instance, Zheng et al. [26] observed that variations in calcium nitrate addition timing during synthesis via the modified St¨ ober method resulted in notable changes in bioactive glass nanoparticle aggregation, composition, and morphology. Similarly, Kesse et al. [27] reported comparable findings using a one-pot synthesis approach, highlighting the critical role of calcium nitrate timing on the particle characteristics such as composition and morphology. Kesse et al. [28] also explored the bioactivity of MBGNPs with high surface area by optimizing the Ca/Si ratio. In this study, the influence of timing of calcium nitrate addition on the structural, textural, and biocompatibility properties of binary MBGNPs synthesized via the microemulsion-assisted sol-gel method was evaluated. The morphology, surface area, and pore structure of MBGNPs were characterized using advanced analytical techniques. Additionally, their biological response was assessed in osteoblast-like cells to evaluate their potential for bone tissue engineering applications. The results were compared with previous studies to further understand how synthesis parameters influence material properties and biological performance. 2. Materials and method 2.1. Synthesize of MBGNPs Tetraethyl orthosilicate (TEOS, 99 %, Sigma-Aldrich), hexadecyltrimethylammonium bromide (CTAB, BioXtra, ≥99 %, H9151, SigmaAldrich), calcium nitrate tetrahydrate (Ca(NO₃)₂⋅4H₂O, 99.1 %, VWR), ethyl acetate (99.7 %, l-00,028, Centralchem), and ammonium hydroxide (ACS reagent, 28.0–30.0 % NH 3 basis, 221,228, Sigma-Aldrich) were used as precursors in the synthesis of MBGNPs with a nominal composition of 70SiO 2 –30CaO (in mol %) via a microemulsion-assisted sol-gel method [29]. In brief, 1.4 g of CTAB was dissolved in 66 ml of deionized water, followed by adding 20 ml of ethyl acetate (EA) and stirring at 25 ◦C for 30 min. Next, 1.8 ml of ammonium hydroxide (28 %) was added and stirred for 15 min. Then, 7.2 ml of TEOS was added, followed by 4.26 g of calcium nitrate tetrahydrate, with varying addition times (15 to 90 min). The mixture was stirred for 4 h at 25 ◦C. Afterwards, the samples were washed twice with deionized water, once with ethanol by centrifugation, and dried overnight at 60 ◦C. The dried samples were calcined at 650 ◦C for 3 h with a heating rate of 1 ◦C/min. The samples were labelled as MBGNP_15min, MBGNP_30min, MBGNP_45min, MBGNP_60min, and MBGNP_90min, corresponding to calcium nitrate addition times of 15, 30, 45, 60, and 90 min, respectively. 2.2. Characterization of MBGNPs The morphology and particle size of MBGNPs were analyzed using field emission scanning electron microscopy (FESEM, Hitachi S4700) at 20 kV, with gold sputter-coating to improve conductivity. Particle size was determined by analyzing at least 100 particles using ImageJ software. The chemical compositions of MBGNPs were analyzed using Wavelength X-ray Fluorescence (WXRF) with a PANalytical MagicX device (PW-2424) equipped with a Rh anode RX tube and a 2.4 kW generator. MBGNPs were homogeneously mixed with anhydrous lithium tetraborate (Li₂B₄O₇) flux and lithium bromide (LiBr) as a release agent. The mixture was melted at 1100 ◦C using a Philips Perl’X3 pearlizer and a platinum-gold crucible. The melted mixture was then cast onto molds of matching composition to produce transparent glass beads. Calibration curves were constructed using certified silico-aluminous standards to correlate the measured X-ray intensities with elemental concentrations. Microstructural properties of MBGNPs were analyzed using transmission electron microscopy (TEM, Talos F200S G2, Thermofisher USA), N 2 adsorption-desorption (ASAP 2020, Micromeritics, USA) and UltraSmall and Small Angle X-Ray Scattering (USAXS/SAXS). For TEM, the samples were dispersed in ethanol via ultrasonic bath, then mounted on carbon grids and imaged at 200 kV. Prior to N 2 adsorption-desorption, MBGNPs were degassed at 150 ◦C for 1200 min. The specific surface area was determined using the Brunauer-Emmett-Teller (BET) method, while the Barrett-Joyner-Halenda (BJH) method was used for pore size distribution analysis. USAXS/SAXS measurements were conducted at 12-ID-E beamline at Advanced Photon Source, Argonne National Laboratory. The combined scattering vector magnitude (q) range was between 1 and 1 ×10 −4 Å −1 ; here q =4 π /λ sin(θ), λ is the wavelength and θ is ½ of the scattering angle. The X-ray energy was 21 keV (λ =0.5895 Å) and the X-ray photon flux density was ≈5 ×10 12 mm −2 ⋅s −1 . Before measurements, the weight of the sample holder was recorded both with and without the MBGNP powders, ensuring the weight of the tape windows was accounted for to facilitate absolute intensity calibration. The volume occupied by the powder was determined from the dimensions of the sample holder. The volume and weight differences allowed the calculation of the packing density of the sample during measurement. The resulting packing density was then used to calibrate the measured signal relative to MBGNP weight. For structural characterization of silicate units, Fourier Transform Infrared (FTIR) spectra were recorded in absorption mode using a Perkin-Elmer Spectrum 3 FTIR Spectrometer (USA) within the 400–4000 cm −1 wavelength range, with a resolution of 4 cm −1 and 32 scans. Magic-Angle Spinning (MAS) NMR spectra were recorded using a Bruker AV-400-WB NMR spectrometer, employing 4 mm ZrO₂ rotors with Kel-F caps. ²⁹Si MAS-NMR spectra were measured at 79.49 MHz with π /2 pulses at 62.5 kHz, a spectral width of 40 kHz, and a MAS speed of 10 kHz. A relaxation delay of 60 s was applied between scans, with kaolin as a secondary reference, calibrated to a chemical shift of 91.2 O. Bas¸ak et al. Open Ceramics 23 (2025) 100807 2 ppm relative to tetramethylsilane (TMS). MAS-NMR spectra were simulated using Dmfit software [30]. 2.3. Ion release behavior of MBGNPs The ion release test was performed in a Tris/HCl buffer (50 mM, pH 7.4). MBGNPs were immersed in the buffer at a concentration of 1.5 mg/ mL and incubated at 37 ◦C with continuous stirring at 120 rpm for up to 7 days. At predetermined time intervals, aliquots were collected, syringe filtered, and supernatants analyzed using inductively coupled plasma optical emission spectroscopy (ICP OES Agilent 5100 SVDV, Agilent Technologies, USA). Calibration solutions were prepared using certified ICP reference standards (Analytika, Prague, Czech Republic), and scandium was employed as an internal standard to minimize nonspectral interferences. Data were reported as mean values ±standard deviations from at least three replicates for each element. 2.4. In-vitro cytotoxicity assay In vitro cytocompatibility was tested using MG-63 human osteoblastlike cells line (Sigma-Aldrich, Germany). The cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, Germany) with 10 % fetal bovine serum (FBS, Gibco, Germany) and 1 % PenicillinStreptomycin (P/S, Gibco, Germany), in a humidified atmosphere with 5 % CO 2 at 37 ◦C. Indirect testing was performed using supernatants. MBGNPs were sterilized by heat treatment at 160 ◦C for 3 h, added to the cell culture medium at 10 % w/v concentration, and then incubated at 37 ◦C for 24 h. Supernatants were collected via centrifugation, filtered with a 0.22 μ m syringe filter, and diluted to 1 %, 0.1 %, and 0.01 % (w/ v) in complete medium for the cytocompatibility assay. MG-63 cells were seeded into 96-well plates at an inoculum density of 5 ×10 4 cells/mL and incubated for 24 h. The cells were then exposed to extracts at 1 %, 0.1 %, and 0.01 % w/v concentrations for 48 h. The samples were prepared in triplicate, and cell viability was assessed using the WST-8 assay (CCK-8 Kit, Sigma-Aldrich, Germany). Cells were incubated with 1 % v/v WST-8 in a cell culture medium for 3 h. A blank Fig. 1. FESEM images and particle size distribution of MBGNPs. O. Bas¸ak et al. Open Ceramics 23 (2025) 100807 3 sample containing only 1 % v/v WST-8 in a cell culture medium was also incubated. After incubation, 100 μ L aliquots from each well were transferred to a new 96-well plate for spectrometric analysis using a microplate reader (NanoEnTek EVE Automatic cell counter) at 450 nm. Relative cell viability was calculated using Eq (1). Statistical analysis was conducted using Origin 2018b software with one-way ANOVA and Bonferroni’s test. Statistically significant differences were represented by p <0.05. The results were presented as mean ±standard deviation (SD). 3. Results and discussion 3.1. Physicochemical properties Fig. 1 presents FESEM images and particle size distribution histograms of MBGNPs, illustrating how the timing of calcium nitrate addition influences nanoparticle morphology and size distribution. The FESEM images reveal that the MBGNP_15min sample exhibits a broad size distribution and predominantly non-spherical, elongated nanoparticles, with the highest aspect ratio (1.44) among all samples. In contrast, other MBGNP samples display a uniform size distribution and spherical morphology, as confirmed by the particle size distribution histogram. In this study, the only variable was the time interval between TEOS and calcium nitrate addition, while all other parameters were held constant. This ensured silica formation began before calcium nitrate was incorporated in all MBGNPs. The non-spherical shape of MBGNP_15min is attributed to insufficient reaction time for complete particle formation. Results indicate that a minimum 30-minute interval between TEOS and calcium nitrate addition is necessary to achieve uniformity in MBGNPs, consistent with findings from microemulsion-assisted sol-gel synthesis studies [10,14,31]. Additionally, the mean particle size of MBGNPs increased with delayed calcium nitrate addition while maintaining dispersity and uniformity, except for MBGNP_15min. A similar trend was observed in studies on MCM-41, which reported a correlation between reaction time and particle growth [32]. The chemical compositions of MBGNPs are summarized in Table 1. The analyzed SiO 2 /CaO ratios are significantly lower than the nominal values (70SiO 2 –30CaO), due to calcium loss during the washing steps employed during synthesis [19]. Nevertheless, the analyzed compositions are consistent with those reported in the existing literature for the binary SiO 2 –CaO system [10,29]. The CaO content in MBGNPs is notably influenced by the amount of calcium nitrate deposited on the silica nanoparticle surfaces. This deposition process is governed by electrostatic interactions between the negatively charged Si–OH groups on silica and Ca 2+ ions [26]. When particle formation is incomplete, silica particles exhibit higher reactivity, enhancing electrostatic interactions. As previously reported, Ca 2+ ions act as cross-linkers between silica particles [27]. Under these conditions, an increased Ca 2+ deposition is observed. However, these ions are prone to aggregation due to their entrapment within the MBGNP structure, complicating their removal during washing steps. This phenomenon explains the higher CaO concentration and potential agglomeration of Ca 2+ ions in MBGNP_15min, which also exhibits an irregular shape. The findings indicate that earlier calcium nitrate addition enhances Ca 2+ deposition, leading to ion entrapment and influencing the chemical composition and morphology of MBGNPs. In contrast, a slight decrease in CaO content is observed with delayed calcium nitrate addition. As shown in Fig. 1, MBGNP_90min exhibits the largest particle size. Larger particles are reported to exhibit greater resistance to electrostatic interactions compared to smaller particles, with the addition of electrolytes such as calcium salts [33]. In the case of structural characterization of silicate units, FTIR and MAS-NMR techniques are employed. The FTIR spectra of MBGNPs (Fig. 2) show three characteristic bands: one at approximately 443–451 cm −1 corresponding to Si–O–Si bending vibrations, another at 803 cm −1 attributed to Si–O symmetric stretching vibrations, and a last broad band between 1000 and 1250 cm −1 assigned as asymmetric stretching. This most intense band has two shoulders due to transverse optical mode (TO, 1050 cm −1 ) and longitudinal optical (LO, 1250 cm −1 ) mode of asymmetric stretching vibration of the Si–O–Si bond in SiO 4 tetrahedra. These spectral features are consistent with those reported for MBGNPs in previous studies [10,29]. The structural units in MBGNPs are described using the Q Si n notation, where "n" denotes the number of bridging oxygens (BO) per SiO 4 tetrahedron, which ranges from 0 to 4 [34]. Q Si 4 units can transform into Q Si 3 , Q Si 2 , Q Si 1 , or Q Si 0 units, promoting the depolymerization of the silicate network structure after CaO addition. This process can be induced by network-modifying cations such as Ca 2+ or H + , which generate non-bridging oxygens (NBOs) within the structure [35]. Furthermore, the 29 Si MAS-NMR spectra were performed on the MBGNP_15min and MBGNP_90min samples, representing end-member compositions due to the consistent CaO content across all MBGNPs. The asymmetric nature of Table 1 Chemical composition of MBGNPs (mol %) measured by WXRF. Samples SiO 2 CaO MBGNP_15min 85.8 (±0.5) 14.2 (±0.1) MBGNP_30min 88.4 (±0.6) 11.6 (±0.2) MBGNP_45min 88.9 (±0.4) 11.1 (±0.2) MBGNP_60min 89.2 (±0.6) 10.8 (±0.2) MBGNP_90min 90.5 (±0.5) 9.5 (±0.1) Fig. 2. FTIR Spectra of MBGNPs as a function of calcium addition time. Cell viability (%) = (Absorbance of sample −Absorbance of blank) (Absorbance of positive control −Absorbance of blank)×100 (1) O. Bas¸ak et al. Open Ceramics 23 (2025) 100807 4 the 29 Si MAS-NMR spectra required a deconvolution process to determine the distribution of Q Si n structural units. Fig. 3 presents the experimental and deconvoluted spectra, while Table 2 provides the corresponding peak assignments, with observed chemical shifts from −110.49 to −110.30 ppm for Q Si 4 and from −101.09 to −100.70 ppm for Q Si 3 units, for 15and 90-minutes samples, respectively. The full width at half maximum (FWHM) values ranged from 11.50 to 11 ppm for Q Si 4 and 8 ppm for Q Si 3 units. The results confirm that the deconvolution method aligns with the protocol established for MBGNPs by Leonova et al. [36]. In this approach, the primary Q Si 4 resonance centered near −110 ppm, substituting hydroxyl groups with BO atoms, resulting in an increase in the chemical shift of approximately 8–12 ppm. The presence of Q Si 3 units in MBGNPs indicates hydroxyl groups and BO atoms within the structure, suggesting that H + cations act as network modifiers like Ca 2+ ions. Fig. 3. 29 Si MAS-NMR Spectra of a) MBGNP_15min and b) MBGNP_90min. Table 2 Peak Assignments in 29 Si MAS-NMR spectra. Sample Q Si 4 Q Si 3 δ (ppm) FWHM (ppm) Population ( %) δ (ppm) FWHM (ppm) Population ( %) MBGNP_15 min −110.15 11 75.30 −100.95 8 24.70 MBGNP_90 min −110.50 11.85 77.54 −101.05 8 22.46 Fig. 4. TEM images of a) MBGNP_15min, b) MBGNP_30min, c) MBGNP_45min, d) MBGNP_60min and e) MBGNP_90min. O. Bas¸ak et al. Open Ceramics 23 (2025) 100807 5 In the MBGNP_15min sample, a higher population of Q Si 3 units than MBGNP_90min suggests a more open silicate network due to the greater incorporation of CaO, as consistent with Table 1. Furthermore, the 29 Si MAS-NMR spectra correlates with the degree of network condensation, where Q Si 4 is the most abundant unit, reflecting rapid silica network condensation [37]. The delayed addition of calcium nitrate increases the Q Si 4 units, indicating a more condensed and highly cross-linked silica network characterized by an increased number of siloxane bridges [38]. This observation is consistent with findings that extended reaction times enhance network condensation [39]. 3.2. Microstructure of MBGNPs Fig. 4 shows the TEM images of MBGNPs, illustrating the influence of calcium nitrate addition timing on the particle morphology and mesoporosity of the MBGNPs. The particles in the MBGNP_15min sample exhibit a more irregular shape than other MBGNPs, which is consistent with the observation from FESEM images. Additionally, the agglomeration of Ca 2+ ions is more pronounced in the MBGNP_15min sample. A notable feature across all MBGNP samples is the presence of alternating grey and pale stripes in the TEM images. This pattern suggests that the particles are formed through a layer-by-layer self-assembly process. This structural characteristic agrees with previous studies on MBGNPs, as referenced in sources [22,24]. The N₂ adsorption-desorption isotherms of MBGNPs are shown in Fig. 5. For MBGNP_15min, the adsorption pattern suggests a Type VI isotherm (Fig. 5, inset), characterized by stepwise multilayer adsorption between 0 and 0.2 P/P o , consistent with the IUPAC classification for non-porous materials [40]. In contrast, other MBGNP samples exhibit Type IV isotherms with Type 3 hysteresis loops, characteristic of mesoporous materials (2–50 nm) and indicative of slit-shaped pores. At low relative pressures (0–0.1 P/P o ), a slight increase in N 2 adsorption suggests the presence of micropores (<2 nm) in all MBGNPs. At high relative pressures (0.9–1.0 P/P o ), N 2 uptake increases proportionally with delayed calcium nitrate addition (MBGNP_45min, MBGNP_60min and MBGNP_90min). This deviation from the expected horizontal plateau for mesoporous materials indicates the presence of macroporosity (>50 nm) in these samples. Consequently, it can be concluded that all MBGNPs, except for MBGNP_15min, exhibit a combination of micro-, meso‑, and macroporosity. The pore size distribution further confirms the mesoporous nature of the MBGNPs, with a primary peak at approximately 3.7 nm and minor peaks between around 4.35 and 12.22 nm. Additionally, a clear proportional relationship was identified between the timing of calcium nitrate addition and the formation of a bimodal pore distribution. The development of larger pores provides a notable advantage by improving the material’s capacity to accommodate and transport larger biomolecules [41]. For example, it has been reported that effective protein loading into mesopores requires pore sizes exceeding 5 nm [42]. As shown in Table 3, all MBGNPs with a delayed calcium nitrate addition exhibit higher surface areas than MBGNP_30min. The specific surface area value of MBGNP_30min are correlated with values reported in the literature [29,43]. Notably, the MBGNP_45min demonstrates the highest surface area among all MBGNP samples, which can be attributed to a more pronounced hysteresis loop between approximately 0.4 and 0.9 P/P o . This feature is indicative of capillary condensation within mesopores. In contrast, the specific surface area of MBGNP_60min is the lowest among all MBGNPs with delayed calcium nitrate addition (MBGNP_45min, MBGNP_60min, and MBGNP_90min). The smaller pore volume value can explain this reduction in surface area observed for MBGNP_60min compared to the other samples. The USAXS/SAXS absolute calibrated intensity profiles of all MBGNPs, analyzed using Irena software [44], reveal three distinct scattering regions across the q-range of 1 Å −1 to 10 −4 Å −1 (Fig. 6a). All profiles are also presented in Fig. 6b. In the high-q region (q >0.02 Å −1 ), the scattering signal originates from nanopores (<2 nm) within individual particles, quantified using Porod’s law under the assumption of spherical pore geometry. The intermediate q-range (0.002–0.02 Å −1 ) corresponds to scattering from primary particle features. At the lowest q-values (q <0.002 Å −1 ), the observed power-law scattering indicates mass fractal arrangements between particle clusters, characteristic of diffusion-limited aggregation during synthesis. The total specific surface area values presented in Table 4 combine contributions from intragranular microporosity (derived from high-q analysis) and interparticle void spaces quantified through medium-q modeling. The spherical pore model applied to high-q data demonstrates a good correlation with N₂ physisorption trends (Fig. 5), validating the methodology for mesoporous systems. MBGNPs synthesized with delayed calcium nitrate addition exhibit systematically higher surface areas (15–20 % increase) compared to MBGNP_30 min, accompanied by reduced fractal dimensionality, suggesting more open architectural arrangements when modifying the precipitation sequence. Notably, MBGNP_15 min required exclusive characterization through USAXS/SAXS measurements due to Fig. 5. a) N 2 adsorption-desorption isotherms and b) pore size distribution (obtained from desorption branch of MBGNPs). Table 3 Textural properties of MBGNPs with N 2 adsorption-desorption method. S BET : Specific surface area, V P : Total pore volume (desorption branch). Sample S BET (m 2 /g) V P (cm 3 /g) MBGNP_30min 466.1 (±2.7) 0.45 MBGNP_45min 746.4 (±5.2) 0.74 MBGNP_60min 515.1 (±8.3) 0.69 MBGNP_90min 698.4 (±3.2) 0.87 O. Bas¸ak et al. Open Ceramics 23 (2025) 100807 6 N₂ adsorption’s inherent limitations in detecting sub-2 nm pores and potential micropore collapse during degassing pretreatment. Despite this methodological divergence, both techniques consistently identify MBGNP_15min as possessing the lowest total porosity (Table 4), indicative of rapid particle aggregation kinetics during its earlier-stage synthesis. The combined USAXS/SAXS profiles and N 2 adsorption data establish that postponing calcium nitrate introduction enhances pore accessibility while moderating fractal cluster density, providing critical structure-property relationships for optimizing MBGNP design. Modeling of regions 2 and 3 in USAXS/SAXS profiles provides valuable insights into the spatial arrangement of particles, with the results summarized in Table 5. The analysis assumes a Gaussian distribution of particle sizes, primarily spherical, except for MBGNP_15 min, where a spheroidal particle distribution was applied. This adjustment was necessary due to the aspect ratio of 1.44, determined from FESEM images (Fig. 1), which indicated slight anisotropy in particle shape. The local arrangement of grains was further characterized by two key parameters: the volume fraction (φ), which reflects the density of particle packing, and the correlation length ( η ), which describes the average distance over which structural ordering persists. For MBGNPs synthesized with delayed calcium nitrate addition, both φ and η values increased compared to earlier-stage additions. This trend indicates that the grains were distributed over a broader spatial range while maintaining localized clustering over longer distances. Specifically, higher φ values suggest closer packing of nearest neighbours, while larger η values imply more extended ordering across the particle network. Together, these findings suggest that delaying calcium nitrate addition enhances both local density and long-range organization within the mesostructure. The compactness of fractal aggregates was evaluated using the parameter P, which reflects the density of fractal clusters. Higher P values indicate reduced empty space within aggregates, corresponding to denser packing. However, no direct correlation between P values and delayed calcium nitrate addition was observed. Instead, the non-linear trend in P values with increased φ values likely arises from the dynamic nature of the O/W emulsion system used during synthesis. In this system, frequent droplet collisions and random Brownian motion influence the distribution of the oil phase, leading to variations in porosity and grain arrangement [23]. These stochastic processes introduce variability into aggregate density, decoupling P from synthesis timing. These findings also highlight the disordered nature of the pore structure in MBGNPs produced with the microemulsion-assisted sol-gel method [43]. The interplay between emulsion dynamics and sol-gel chemistry creates metastable arrangements that are reflected in the structural parameters derived from USAXS/SAXS modelling. A schematic 3D representation of the modeling is proposed in Figure S1 of the Supplementary Information. To understand the underlying mechanism of the delayed addition of calcium nitrate, it is essential to emphasize that the total reaction time has a more significant influence on the properties of MBGNPs than the amount of CaO incorporated. This is evident from the substantial increase in surface area and pore volume without notable changes in chemical composition. In a basic medium, porosity is primarily governed by electrostatic interactions between positively charged surfactants (S + ) and negatively charged silica clusters (I − ) [45]. However, the presence of counterions (M + X − ), such as salts, can interfere with these electrostatic interactions, potentially altering the porosity. Recent studies suggest that high porosity is primarily attributed to favorable electrostatic interactions between CTAB and silica in a basic medium (pH >8), particularly for larger particles (>100 nm) where minimal pH Fig. 6. (a) Schematic representation of distinct regions observed in the USAXS/SAXS profiles. (b) USAXS/SAXS profiles of all MBGNP samples. Table 4 Specific surface area values of MBGNPs determined by the USAXS/SAXS method. Sample S total (m 2 /g) S fine (m 2 /g) Pore Volume (cm 3 /g) MBGNP_15min 27 21.4 0.02 MBGNP_30min 34 32.5 0.03 MBGNP_45min 144 130.3 0.13 MBGNP_60min 125 107.2 0.15 MBGNP_90min 109 105.3 0.12 Table 5 Fitting parameters of regions 2 and 3 from USAXS/SAXS profiles. Sample Diameter (nm) Width (nm) P (unitless) η (nm) φ MBGNP_15 min 151 22 2.34 84*1.36 * MBGNP_30 min 175 26 1.90 126.5 2.62 MBGNP_45 min 185 39 1.80 180 3.50 MBGNP_60 min 180 38 1.63 225 3.44 MBGNP_90 min 175 37 1.92 255 4.00 * =not reliable. O. Bas¸ak et al. Open Ceramics 23 (2025) 100807 7 drop occurs over time [46]. These interactions are critical for mesophase formation. However, incorporating Ca 2+ ions can reduce these electrostatic interactions, potentially acting as a “blocking effect” that impedes mesophase development [47]. This suppression effect can be correlated with TEM images (Fig. 4) and N 2 adsorption-desorption results (Fig. 5). Notably, MBGNP_15min exhibits incomplete desorption behavior, classifying it as a non-porous material based on N 2 adsorption-desorption isotherms. Furthermore, USAXS/SAXS results confirm that MBGNP_15min has the lowest porosity among all MBGNP samples. In contrast, delayed calcium nitrate addition (MBGNP_45min, MBGNP_60min, and MBGNP_90min) results in more pronounced hysteresis loops in the N 2 adsorption-desorption isotherms. These loops indicate capillary condensation within mesopores and contribute to an enhanced specific surface area compared to MBGNP_30min. USAXS/- SAXS profiles confirm that the delayed addition of calcium nitrate leads to an increase in fine porosities, resulting in higher specific surface area values that are consistent with N 2 adsorption-desorption isotherms. This observation also suggests that Ca 2+ ions hinder mesophase formation, which correlates with the increased S fine values in MBGNPs with delayed calcium nitrate addition compared to MBGNP_30min. Furthermore, delayed calcium nitrate addition leads to the formation a bimodal pore distribution, as evidenced by the coexistence of mesopores and macropores, which is observed with N 2 adsorption-desorption isotherms. Macropores become increasingly apparent with delayed calcium nitrate addition, as demonstrated in Fig. 1. This observation aligns with the broader pore size distribution observed (Fig. 5). The pronounced formation of macropores can be correlated with the USAXS/SAXS profiles in regions 2 and 3, as well as with the N 2 adsorption-desorption isotherms at high relative pressures. However, a discrepancy exists between the specific surface area values of MBGNP_60min and MBGNP_90min, as the results from USAXS/SAXS profiles and N 2 adsorption-desorption isotherms do not fully correlate. The higher specific surface area obtained from N 2 adsorption-desorption for MBGNP_90min compared to MBGNP_60min (Table 3) may be attributed to a greater proportion of fine porosities, as indicated by the USAXS/- SAXS measurements (Table 5). These findings further support that MBGNPs exhibit bimodal porosity, with the dynamic nature of O/W emulsions being more distinctly observed in USAXS/SAXS measurements. 3.3. Ion release behavior in Tris/HCl buffer The ion release behavior of nanoparticles was studied using Tris/HCl buffer at pH 7.4, as shown in Fig. 7. Among the samples, MBGNP_15min exhibited the highest release of Ca 2+ and Si 4+ after 1 day of immersion. Interestingly, the concentration of Si 4+ is similar after 3 and 7 days of incubation. The ion release behavior of MBGNPs is primarily influenced by two factors: specific surface area and chemical composition [48]. However, its chemical composition (Table 1) shows the highest calcium oxide content (14 mol%), and 29 Si MAS-NMR results indicate a greater number of NBOs. These factors likely contribute to its accelerated ion release. The MBGNP_30min and MBGNP_45min samples exhibited similar ion release behavior for both Ca 2+ and Si 4+ ions. Specifically, MBGNP_30min showed higher Si⁴⁺ release after 1 day of incubation, but after 3 and 7 days, the silicon concentrations became comparable. For Ca 2+ release, both samples demonstrated similar levels after 1 and 3 days, with MBGNP_30min showing a slightly higher release on day 7. In contrast, the MBGNP_60 min and MBGNP_90min samples exhibited significantly lower ion release for both ions than the other samples. After 7 days, both samples released similar concentrations of calcium ions. However, MBGNP_90min showed a slower release rate. A similar trend was observed for Si 4+ ions, with MBGNP_90min showing an approximately 25 % slower release rate than MBGNP_15 min. The combined influence of chemical composition and specific surface area on ion release behavior presents challenges for explicit interpretation due to their simultaneous variation [49]. The effect of chemical composition is particularly evident when comparing the ion release rates of MBGNP_15min and MBGNP_90min. As shown in Table 3, the MBGNP_15min has a higher number of NBOs and a less connected silica network than MBGNP_90min, leading to accelerated dissolution despite its lower surface area. This observation is consistent with findings reported by Schumacher et al. [50]. Conversely, the impact of specific surface area is more pronounced when comparing MBGNP_45min and MBGNP_60min. Despite their nearly identical chemical compositions (Table 1), the faster ion release observed for MBGNP_45min can be attributed to its higher specific surface area [51]. While MBGNP_30min is a reference point in this study, notable differences in ion release behavior are evident across the MBGNPs. Despite delayed calcium nitrate addition resulting in higher specific surface areas (MBGNP_45min, MBGNP_60min and MBGNP_90min), the fastest ion release among these samples was observed for MBGNP_30min. Notably, the highest ion release was observed for MBGNP_15min. This finding highlights that surface reactivity plays a crucial role in ion release behaviour alongside specific surface area and chemical composition. The elevated release rates observed for MBGNPs, such as MBGNP_15min and MBGNP_30min, can be attributed to their higher levels and more homogeneous distribution of CaO on particles, as also confirmed by USAXS/SAXS profiles. Furthermore, Tris/HCl buffer maintains physiological pH conditions due to its buffering capacity; however, the ion release behaviour of nanoparticles may differ under real-life conditions, where the Fig. 7. Ion release profiles of MBGNPs for a) Ca 2+ (mg/L) and b) Si 4+ (mg/L) ions in TRIS/HCl Buffer solution. O. Bas¸ak et al. Open Ceramics 23 (2025) 100807 8 environment may promote precipitate formation or protein adsorption on the material’s surface [52,53]. Therefore, it is important to conduct ion release studies in various media, such as simulated body fluid or cell culture medium, to better understand the interactions of the nanoparticles. Future studies should also evaluate the long-term ion release kinetics in dynamic biological environments to more accurately simulate in-vivo conditions and assess potential effects on bioactivity and biocompatibility. 3.4. In-vitro biocompatibility Fig. 8 presents the cell viability of MG-63 osteoblast-like cells cultured for 48 h with diluted extracts of MBGNPs. Based on the relative cell viability, all samples demonstrated non-cytotoxic behavior in accordance with the ISO 10,993–5 standard, which defines materials with cell viability above 80 % as non-cytotoxic. The lowest cell viability was observed at the 1 % w/v extract concentration for all MBGNPs. These findings are consistent with the study by Kurtuldu et al. [29], which evaluated the cytocompatibility of MBGNPs extracts based on a binary SiO 2 –CaO system. Their results indicated promising cell viability at extract concentrations of 1 % w/v and lower. For samples MBGNP_15min, MBGNP_30min, and MBGNP_45min, an increase in cell viability was observed with decreasing extract concentrations. This outcome highlights the anticipated role of Ca 2+ ions in enhancing osteogenic cell proliferation, such as MG-63, while the release of Si 4+ ions further enhance cell viability [54]. As shown in Fig. 6, the most reactive material is MBGNP_15min, which exhibited the fastest ion release during immersion in TRIS/HCl buffer solution. For MBGNP_45min, the highest cell proliferation occurred at an extract concentration of 0.1 % w/v, suggesting that this concentration provides an optimal microenvironment for osteogenic cells, likely due to the balanced therapeutic ion release. Conversely, the lowest cell viability was observed for MBGNP_60min and MBGNP_90min, which exhibited the slowest ion release rates. This reduced viability may be attributed to insufficient release of Ca 2+ and Si 4+ ions, essential for fostering a stimulatory environment for osteogenic cells. Suboptimal ion release likely results in less favorable conditions for cell proliferation [55]. Overall, MBGNPs with faster ion release (MBGNP_15min, MBGNP_30min, and MBGNP_45min) more effectively support osteogenic cell proliferation by providing adequate levels of therapeutic ions. In contrast, those with slower ion release (MBGNP_60min and MBGNP_90min) exhibit slightly reduced cell viability, likely due to insufficient Ca 2+ and Si 4+ ion release, which diminishes their stimulatory effects on osteogenic cells. 4. Conclusion This study demonstrates that the timing of calcium nitrate addition considerably impacts the porosity of MBGNPs. Delayed addition of calcium nitrate increased specific surface area, pore size, volume, and particle size. While the discrepancy between nominal and analyzed compositions was primarily attributed to the washing steps, the timing of calcium nitrate addition also influenced the chemical composition due to particle enlargement and larger pore sizes. Despite this, the proportion of Q Si n units slightly changed, indicating minimal variation in the chemical composition across the samples. The results of the study indicated that the presence of any degree of toxicity in osteoblast-like cells was not observed among all MBGNPs. These findings suggest that adjusting the timing of calcium nitrate addition is a simple yet effective strategy for tailoring the textural properties of MBGNPs for various biomedical applications. Additionally, a 45-minute interval between the addition of TEOS and calcium nitrate seems optimal to produce highquality MBGNPs, offering a valuable guideline for future research. CRediT authorship contribution statement Onat Bas¸ak: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Fatih Kurtuldu: Writing – review & editing, Visualization, Methodology, Investigation, Formal analysis, Conceptualization. Jan Ilavský: Writing – review & editing, Visualization, Investigation. Martina Vit´ azkov´ a: Writing – review & editing, Investigation. Ana M. Beltr´ an: Writing – review & editing, Investigation. Francisco Mu˜ noz: Writing – review & editing, Methodology, Investigation. Yolanda Castro: Writing – review & editing, Supervision, Resources, Investigation. Martin Mich´ alek: Writing – review & editing, Supervision, Resources. Maria Chromˇ cíkov´ a: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Fig. 8. Cell Viability results of MG-63 cells cultured with the extract of 1, 0.1 and 0.01 % cell culture medium dilutions (*p <0.05). O. Bas¸ak et al. Open Ceramics 23 (2025) 100807 9